Study On Quality And Safety Discrimination Of Mutton By Low-field Nuclear Magnetic Resonance And Electronic Tongue Technology

In recent years, with the development of national economy and improvement of living standards, our country has become mutton production and consumption power. But in the current, to put on weight by means of water-injection and gum-injcetion has been occured more frequently. Adulteration of meat, involving the replacement of selected breeds, particular traditional method with other cheaper animal proteins, has jeopardized the market r egulation and consumers’ health. This paper focuses on application of adulteration and freshness detecting by Low-field NMR and Electronic tongue technology. To explore the application of non-destructive testing technology based on low-field NMR and Electronic tongue technology in mutton quality and safety.Through two kinds of testing instruments, combined with a variety of pattern recognition methods for data analysis, the best method of pattern recognition and prediction model was established.Using LF-NMR technology on normal mutton and water injection, injection plastic mutton detection, according to the sample of T2 time reflects water existence state and the distribution of results, combined with principal component analysis(PCA) and stepwise discriminant analysis(Step-LDA) to distinguish between different samples of mutton identification and cross validation. The results shows that based on the relaxation properties of the sample combined with PCA method; Pure mutton and water-injection,gum-injection mutton can effectively distinguish, different proportions of water-injection, gum-injection mutton also can be distinguish.Step-LDA of the distinction between these samples was more significant than PCA, the samples on discriminant score chart presents regular distribution along with the change of adulteration ratio. Through cross validation, the classification accuracy achieve above 100%, 70% and 100% respectively. Discriminant model was better than that of the principal component analysis. LF-NMR technology in combination with the appropriate data algorithm has great application potential which in the mutton quality control and evaluation.Using electronic tongue to detect adulterated samples mixed with duck, pork and other less expensive meats. It can be seen on the radar map of taste, add other substances would cause the five sensors of E-tongue change in varying degrees. Different proportions of adulterated samples showed a regular distribution along the axis, and has a linear relationship. Principal component analysis showed that the sample of pure mutton and different types of adulteration can be respective distinguish. Discriminant analysis model constructed by pure and adulterated mutton can effectively distinguish adulteration and pure mutton. The cross-validation correctly distinguish rate of 100%. This chapter provides a fast and accurate new method for identification of adulterated meat.To detect different storage time of mutton through physical and chemical experiments, and microbiological test.Study on the changing of p H, TVB-N, the fleshcolor and the total number of colonies during storage of mutton. The results shows that there are regular changes and a significant correlation in the indicators during the storage period. Using cluster analysis combined with p Hysical and chemical indicators to reflect the changing of quality during storage,then grading the mutton. The results of grading shows that the sample of 1-4 days for freshness, 6-8 days for the deterioration of the critical time, eight days later for corruption meat.Application of LF-NMR techniques to study the distribution and variation of moisture during storage, and application of electronic tongue technique to study the changes during storage mutton taste.Using LF-NMR and E-tongue to detectthe freshness of mutton was feasible. Freshness rating prediction models suggest that LF- NMR has significant correlation with electronic tongue between technology and conventional indicators characterizing freshness. By LF- NMR detection and multivariate statistical analysis, can accurately distinguish between different levels of freshness mutton. Statistical analysis combined with relaxation properties shows that the representative data of freshness was T22(relaxation time of unease flow water), T23(relaxation time of free water), P22(peak area ratio of unease flow water), P23(peak area ratio of free water). Taste-dimensional radar map and chart shows that the most significant change during mutton storage was AAE(umami) and CAO(sour),which can be used to evaluate the freshness of the mutton. Predictive model of mutton freshness, established through Fisher discriminant analysis was 90.09% accurate, the prediction accuracy was above 92.31%.The model predicting the freshness changing rates of E-tongue based on taste was established. The result showed the samples of different freshness could be well distinguished. The predicting model of freshness was feasible(accuracy=100%). The results shows that the mutton freshness can be predicted correctly by the model based on taste. The model was stable and has broad prospects in the production practice.